It is three in the afternoon in late July. The thermostat in the apartment reads 94 degrees. The box fan in the window has been running since noon, and it feels like the responsible thing to do.
It is not. The Centers for Disease Control and Prevention tells people to “Use fans, but only if indoor temperatures are less than 90°F.” Above that, CDC says, a fan can increase your body temperature.
Almost nobody knows this. Fans are the one piece of heat advice everybody thinks they already understand, and in the conditions that kill people, the advice reverses. That single correction is worth more than everything else in most heat-safety articles, so it goes first.
What follows covers who gets hurt, how to tell a bad afternoon from a 911 call, why your prescriptions belong in a heat plan, and the half of this subject nobody covers: cold, which does its damage in American homes at temperatures that sound mild.
Why a fan flips from help to hazard around 90°F
A fan does two things to your body, and only one of them is cooling.
The useful one is evaporation. Moving air carries away the water on your skin, and evaporating water takes heat with it. That is the main way a human being sheds heat when the surroundings are warm.
The other is plain heat transfer. Air moving across your skin exchanges heat in whichever direction the gradient runs. When the air is cooler than your skin, that cools you. When the air is hotter than your skin, a fan delivers heat to you, the way a convection oven cooks faster than a still one. At some point the heat the fan brings in overtakes the cooling the evaporation provides, and CDC’s 90°F cutoff is where it draws that line for a room.
Researchers have measured this directly. A 2025 randomized trial published in the Journal of Applied Physiology put 23 adults aged 66 to 84 through three hours of heat at 47°C, about 117°F, at 15 percent relative humidity, while they did ordinary activities. Each person went through four conditions: water sprayed on the skin, an electric fan, both, or nothing. Fan use increased the rise in core body temperature by 0.60°C compared with no cooling at all, raised the ending heart rate by 15 beats per minute, and pushed sweat rates up. Water spray went the other way, trimming the core temperature rise by 0.24°C. The authors concluded that electric fans are not advised as a cooling strategy in those conditions.
The honest version has limits. Those were older adults in very hot, very dry air, not an ordinary muggy American afternoon. Humidity changes the arithmetic, since humid air slows evaporation and dry air speeds it. Where the crossover sits across the full range of heat and humidity is still being worked out, which is part of why CDC gives a plain temperature cutoff instead of a formula.
The trade-off, stated honestly: wetting your skin first changes the picture. In that same trial, spray plus fan did not raise core temperature, so the fan stopped being actively harmful once the skin was wet. Heart rate still ran higher than with spray alone. Wet skin plus a breeze beats a bare fan in hot air. Neither replaces a cooler room.
The practical rule fits in a sentence. Below 90°F indoors, a fan earns its place. Above 90°F, the fan is not your cooling plan.
Who heat reaches first
CDC organizes its risk list by who needs extra action on hot days, and it is worth regrouping by why.
Some bodies regulate temperature less well. CDC names people 65 and older, and infants and young children, who have less reserve in both directions. Article 676 covers home changes for aging in place, including temperature extremes in later life.
Some conditions are made worse by the strain. CDC specifically names people with heart disease, and people with asthma, especially children. Pregnancy is on the list too. The agency keeps a wider index of heat risk factors, because the list of chronic conditions that interact with heat is long.
And some people are exposed by what they do. People working or exercising outdoors take on heat they cannot step away from.
Then CDC says the thing that keeps this from being somebody else’s problem. Anyone, at any time, can have a heat-related illness. Healthy people in their twenties go down in heat, and the usual reason is acclimatization, the physical adaptation the body makes over several days of repeated heat exposure.
The evidence there is specific. A 2018 analysis in the Journal of Occupational and Environmental Medicine reviewed 38 heat illness enforcement investigations by the Occupational Safety and Health Administration covering 66 cases, and found 73 percent of the fatal cases happened during the worker’s first week on the job. The first week is when the body has not yet adapted. The same logic covers a teenager starting a landscaping job and anyone back from two weeks in an air-conditioned office.
Heat exhaustion, heat stroke, and the line that decides whether you dial
CDC’s main heat page lists symptoms as one group: muscle cramping, unusually heavy sweating, shortness of breath, dizziness, headaches, weakness, and nausea. It tells readers to know when to seek medical care. It does not draw the line between heat exhaustion and heat stroke, so this section is sourced to a page that does, the National Institute for Occupational Safety and Health’s page on heat-related illnesses.
NIOSH describes heat exhaustion with headache, nausea, dizziness, weakness, irritability, thirst, heavy sweating, reduced urine output, and a body temperature that is elevated without a specific number attached. Its guidance is to get the person out of the heat, loosen or remove extra clothing, cool the head, face and neck with cold compresses, give frequent small sips of cool water, and get them evaluated at a clinic or emergency department, calling 911 if care is not otherwise available.
Heat stroke is a different animal. NIOSH lists confusion, slurred speech, seizures, and loss of consciousness, and says body temperature can reach 106°F or higher within 10 to 15 minutes. Its instruction is unambiguous: call 911.
Heat stroke is a medical emergency. Call 911. Confusion, loss of consciousness, or hot dry skin in someone who has stopped sweating are not wait-and-see signs.
One detail in that list corrects a widespread half-truth. People are taught that hot, dry skin means heat stroke and sweating means you are still in the safer category. NIOSH lists the skin finding as hot and dry or profusely sweating. Somebody drenched in sweat can be in heat stroke. The reliable sign is the brain, not the skin. Confusion, slurred speech, strange behavior, or not tracking a simple question is what should move your hand to the phone.
While you wait, cooling matters and speed matters. NIOSH’s first aid for heat stroke is to move the person to shade, remove outer clothing, and cool fast, with cold water immersion if that is possible, or by wetting the skin, soaking their clothing, and putting cold packs on the head, neck, armpits and groin. A 2026 review in American Family Physician treats rapid cooling, ideally immersion within the first 30 minutes, as central to how heat stroke turns out.
Your medicine cabinet belongs in your heat plan
This is the part CDC covers and almost no consumer article mentions.
CDC makes three separate points about medicines and heat, and they do different jobs. The first is that many medicines can leave you dehydrated or overheated on hot days. CDC’s clinical guidance for health care providers is blunter: many medications can impair heat tolerance and predispose people to heat illness.
The mechanism is not mysterious. A 2020 review in the International Journal of Women’s Dermatology lays out that the skin is the organ tasked with dumping heat, which it does by expanding blood flow to the surface and by sweating. Drugs that blunt sweating, reduce the heart’s ability to raise output, or alter fluid balance interfere with exactly that machinery.
The second point is the one people get wrong in the other direction. Do not stop or change a medicine because of the weather. CDC says to talk to your doctor first, and that is not a formality. The condition the drug treats does not pause for a heat wave, and some of these medicines are more dangerous to stop than to continue.
The third is storage and power. Heat causes outages, and CDC tells people to have a plan for refrigerated medications and electronic medical devices before one happens. Some medicines also need to be kept out of hot places, which includes a mailbox, a car, and a porch where a delivery sits for six hours in August.
Here is a short list of questions worth taking to a pharmacist once, before summer, rather than during a heat wave:
- Does anything I take affect sweating, thirst, blood pressure, or how my body handles heat?
- Which of these have a temperature limit, and what should I do about mail-order deliveries in summer?
- Which need refrigeration, and roughly how long do they hold in a cooler if the power fails?
- If I have been told to limit fluids for a heart or kidney reason, what changes during a heat wave?
- Is there anything here I should ask my prescriber about specifically before hot weather?
That conversation takes ten minutes and does not need an appointment.
Cooling a home that does not have air conditioning
Start with the building, during the day. Block the sun on whichever side is getting it, because glass is where most of the heat arrives. Keep the house closed while outdoor air is hotter than indoor air, and open it up once that reverses, usually overnight. Use kitchen and bathroom exhaust fans to push hot air out rather than stir it around. Article 685 covers ventilation in detail.
One reversal is worth flagging. On smoky days, the advice to open the windows at night flips, because what comes in with the cooler air is wildfire smoke. Article 694 handles that conflict.
Then cool the person, not the house. This is where the water spray result from that 2025 trial becomes practical advice: a cool shower, a wet cloth on the neck, a spray bottle, or feet in a basin of cool water does more in a hot apartment than a fan does. Cooling one room well beats cooling a whole house badly.
When the home cannot be made safe, leave it. CDC’s direction is to find air conditioning, and it names three ways: dial 2-1-1, check the National Center for Healthy Housing’s list of cooling centers by state, or call your local health department or emergency management agency. A library, a shopping center, or a relative’s spare room all count.
For planning rather than reacting, CDC’s HeatRisk Dashboard takes a zip code and returns a local heat risk forecast, which lets you see a bad stretch coming instead of discovering it on day two.
Outages deserve their own thought in both directions. A heat wave strains the grid; a winter storm takes lines down. Anybody who depends on an oxygen concentrator, a CPAP, a powered wheelchair, or refrigerated medicine needs a plan written while the lights are on.
Cold is the half everyone forgets
Heat gets the coverage. Cold does damage in American homes at temperatures most people would not call dangerous.
CDC states that hypothermia, meaning abnormally low body temperature, can happen even at cool temperatures above 40°F when somebody is chilled by rain, sweat, or cold water. Its at-risk list is domestic, not alpine: older adults with inadequate food, clothing or heating, and babies sleeping in cold bedrooms.
The warning signs in adults are shivering, exhaustion, confusion, fumbling hands, memory loss, slurred speech, and drowsiness. In babies, CDC names bright red, cold skin and very low energy. If a thermometer reads below 95°F, CDC says to get medical attention immediately, and it calls hypothermia a medical emergency. Notice the overlap with heat stroke: confusion again, in both directions. Altered thinking is the common alarm.
For a minimum indoor temperature, the clearest published figure comes from the World Health Organization’s Housing and Health Guidelines, published in 2018. For temperate and colder climates in cold seasons, WHO proposes 18°C, about 64°F, as a safe and well-balanced indoor temperature for general populations, with a higher minimum for older people, children, and people with chronic illness, particularly cardiorespiratory disease. WHO grades the recommendation as strong and the supporting evidence as moderate. No US federal agency publishes a comparable household minimum in the heat and cold guidance reviewed for this article, so WHO’s number is the usable benchmark.
Report that at its real strength. A 2023 systematic review in Public Health screened 1,209 studies, included 20, and found that 17 of them linked indoor temperatures under 18°C to worse outcomes across cardiovascular, respiratory, sleep and general health measures. Its own conclusion was that the evidence is not yet sufficient to set specific temperature thresholds for specific population groups. So 18°C is a sensible floor to aim at, not a clinical cut point where something switches on.
The mechanism behind winter excess deaths is mostly blood pressure. A 2019 study in Hypertension measured home blood pressure and indoor temperature in roughly 2,900 people across Japanese households and found morning systolic pressure rose about 8.2 mmHg for every 10°C drop in indoor temperature, with older residents and women more sensitive. Those exact figures come from Japanese housing and should not be transplanted, but the direction is consistent across this literature.
Then there is the reason many American homes are cold, which is money. Preliminary 2024 data from the Residential Energy Consumption Survey, run by the U.S. Energy Information Administration, covers 132.5 million households. Of those, 32.9 million reported reducing or going without food or medicine in order to pay an energy bill. Another 17.6 million reported keeping their home at an unhealthy temperature. About 16.2 million received a disconnect or delivery stop notice, 6.8 million could not use their heating equipment, and 8.2 million could not use air conditioning.
That first figure is roughly one household in four. Safety advice that assumes a working thermostat and an affordable bill is written for about three quarters of the country.
Humidity changes what the same number feels like
Sweat only cools you when it evaporates, and humid air slows evaporation down. That is why a muggy 88°F day can stress the body harder than a dry 100°F one, and why two people reading the same thermometer in different states are not in the same situation.
It also explains a limit mentioned earlier. The fan trial used air at 15 percent relative humidity, which is bone dry, so its numbers do not transfer directly to a humid afternoon on the Gulf Coast. Article 723 covers indoor humidity itself, including the devices sold to change it.
What to set up before the next hot week
Learn two numbers and one sign.
The first number is your indoor temperature, read off a thermometer rather than guessed from how the room feels. Ninety degrees is the fan line, and most households have never checked which side of it they are on during an afternoon.
The second number is in the forecast. Put your zip code into CDC’s HeatRisk Dashboard before a hot stretch rather than during it, and decide in advance which day you will spend somewhere cooler.
The sign is confusion. Every statistic in this subject sits behind one practical moment: somebody in a hot room, or a cold one, who stops making sense. If that happens, stop managing the room and call 911.
One more thing that belongs on the same list. Heat death counts are not on CDC’s main heat page, and the numbers that do exist understate the problem. The Environmental Protection Agency’s climate indicator on heat-related deaths, drawing on CDC WONDER death certificate data, reports more than 14,000 US deaths since 1979 with heat as the underlying cause, record rates in 2021 and 2022, and roughly 750 deaths a year when contributing causes are counted, against more than 1,300 a year under some statistical approaches. EPA says plainly that “this indicator very likely underestimates the number of deaths caused by exposure to heat,” because death certificates often record a cardiovascular or respiratory cause instead. When you read that heat killed a few hundred Americans last year, you are reading the floor, not the number.
Sources
- Centers for Disease Control and Prevention. “About Heat and Your Health.” cdc.gov
- National Institute for Occupational Safety and Health (CDC). “Heat-Related Illnesses.” cdc.gov
- Centers for Disease Control and Prevention. “Clinical Overview (Heat and Health, Health Care Providers).” cdc.gov
- Centers for Disease Control and Prevention. “Preventing Hypothermia.” cdc.gov
- U.S. Environmental Protection Agency. “Climate Change Indicators: Heat-Related Deaths.”
- World Health Organization. “WHO Housing and Health Guidelines, Chapter 4: Low Indoor Temperatures and Insulation.” ncbi.nlm.nih.gov
- U.S. Energy Information Administration. “Household Energy Insecurity, Table HC11.1, 2024 Residential Energy Consumption Survey.” eia.gov
- Journal of Applied Physiology. “McKenna ZJ et al. Evaluating low-energy cooling strategies on thermal and cardiac strain in older adults exposed to very hot and dry heat with accompanying activities of daily living. 2025. PubMed. DOI: 10.1152/japplphysiol.00390.2025.” doi.org
- Public Health. “Janssen H et al. Cold indoor temperatures and their association with health and well-being: a systematic literature review. 2023. PubMed. DOI: 10.1016/j.puhe.2023.09.006.” doi.org
- Journal of Occupational and Environmental Medicine. “Tustin AW et al. Risk Factors for Heat-Related Illness in U.S. Workers: An OSHA Case Series. 2018. PubMed. DOI: 10.1097/JOM.0000000000001365.” doi.org
- Hypertension. “Umishio W et al. Cross-Sectional Analysis of the Relationship Between Home Blood Pressure and Indoor Temperature in Winter: A Nationwide Smart Wellness Housing Survey in Japan. 2019. PubMed. DOI: 10.1161/HYPERTENSIONAHA.119.12914.” doi.org
- International Journal of Women's Dermatology. “Williams ML. Global warming, heat-related illnesses, and the dermatologist. 2020. PubMed. DOI: 10.1016/j.ijwd.2020.08.007.” doi.org
- American Family Physician. “Gauer RL, McNutt R, Bryan K. Heat-Related Illnesses. 2026. PubMed. PMID: 42101601 (no DOI listed in the PubMed record).” pubmed.ncbi.nlm.nih.gov